&   SEARCH
David A. Cushman logo
Cell Size, By Philip Denwood

This article is reproduced here with the author's permission, as he was also the editor of the magazine 'Bee Improvement and Conservation' in the Autumn 2003 edition of which it first appeared, permission is implied by the BIBBA organ as well.

The article is contained within the white panel below and has not been altered in any way, other than being laid out as a web page rather than the columnar form in which it was published. I am grateful to Philip for providing the original text and pictures. Editors comments in square brackets are those of Philip as the 'original' editor, not myself.

   Cell Size  Philip Denwood   

The first experiments in making comb foundation seem to date from the 1840s and 50s. Ever since its commercial manufacture was perfected in the decades that followed, most beekeepers have bought foundation on the quality, price and service offered by the manufacturer, without giving the size of the resulting worker cells a second thought. In certain circles, however, debate on the subject has proceeded at various times. The so-called "large cell controversy" sparked off by the Belgian scientist Ursmar Baudoux rumbled on and off between the 1890s and 1960s, and was revived from a different angle by Beowulf Cooper and his collaborators in the 1960s and 70s. Recently the activities of the Arizona beekeepers Ed and Dee Lusby have set off what may come to be known as a "small cell controversy".

Many beekeepers and scientists before and after the invention of foundation had measured the size of worker cells in natural comb (which, as they often pointed out, tends to vary even within the same colony). Figures ranging from about 750 to about 950 cells per square decimetre (counting both sides of the comb) were arrived at. One of the pioneers of commercial foundation manufacture, the American A.I. Root, standardised in the 1870s on 825-850 as a result of experiment and observation of what the bees he was using seemed to build out most readily. According to H.C. Dadant, referring to American practice in the mid 20th century, "standard comb foundation usually is made on dies providing 857 cells per square decimetre, but the resulting cells may vary from this and usually are larger due to stretching of the wax in manufacture." This would probably bring the size within the range used by Root.

Samples of recently produced foundation have been measured as follows (figures are approximate for reasons to be given):

South Africa (1 sample)     1050 cells per dm2
USA (4)1000-838
New Zealand (2)875-790
Denmark (1)800
France (1)780
UK (4)795-760

The large cell controversy

In an article in the Bee World of 1933, Prof. Ursmar Baudoux recalls the origin of the large cell controversy:

"About 1891, foundation with cells 920 to the square decimetre was introduced into our country [Belgium]. Beekeepers all adopted this size of cell. The experts of that time believed that it was advantageous to produce as many bees as possible on the least possible surface of comb. Thus there was a premature narrowing of the cells, and at the end of a few years the bees were miserable specimens."

"It was then that, to combat so harmful a tendency, I published an article in Progrès Apicole (June 1893) advocating the use of larger cells, as a result of experiments duly described. I had experimented up to the limit of 750 cells per sq. dm."

In later years Baudoux was to standardise on 700 cell foundation, and experimented with as low as 675, in which the floor area of a cell was 26.6% larger than in 920 foundation. His guiding principle was, in his own words, "The size of the bee is correlated with the size of the cell. Small cell, small bee; big cell, big bee." Colonies of the bees reared in the larger cells and established as larger themselves by many measurements were alleged to produce significantly more honey than colonies of smaller bees under otherwise identical conditions.

He made no bones about the fact that he was aiming primarily at higher honey production:
"The end proposed is magnificent: To rear bees of extraordinary vigour, able to forage over a more extended flight-radius and to visit a multitude of flowers the nectar of which is, at present, out of reach of their tongues. The tongue of the bee is, indeed, the essential organ which we must develop, by selection first, and then by rearing the breed in large cells; for it is plain that all the organs will participate in any enlargement of the body of the bee." [Italics original]
Baudoux was an experienced beekeeper who made extensive measurements to back up his theories, but he had his detractors as well as his supporters. He was accused of adhering to the Lamarckian theory of evolution which claimed that characters acquired in one lifetime could be inherited by subsequent generations: a theory which was eventually discredited in favour of the Darwin/Mendel one. Even if the accusation is true it need not invalidate the theory that large cells produce large bees in a given generation. Also a peculiarity of the way bees produce comb may give a pseudo-Lamarckian effect, as I shall suggest below.

The controversy was followed up by researchers in many countries, most of whom concluded that the size of cells does indeed affect the size of the bees reared in them, though not necessarily to the degree that Baudoux had claimed. Much of the research was reviewed by Roy A. Grout who also made extensive measurements for his M.Sc. Thesis. He concluded in 1931 (among other things) that:
"1. The size of the worker bee as represented by the size of the various parts is significantly increased through the use of brood combs containing enlarged cells.

2. The average percent of increase of the linear measurements of the worker bee is directly proportional to the percent of increase of the diameter of the brood cell.

3. The number of bees used in a sample in this experiment is not large enough to give wholly consistent results, but these results are in general significant and indicative."
On the question of whether, other things being equal, big bees are better than small ones, authorities continued to be divided, as indeed they still are. Some of the opinions on both sides were backed up by extensive experiments. Already before 1910 a Frenchman called Pincot reported independently of Baudoux that over two years, 30 colonies with large cells gathered 30% more honey than others. In 1965 C. Antonescu in Romania concluded after over 20 years of tests that:
"Experiments with a large cell honey-comb in the conditions of the Socialist Republic of Romania (5.65 mm [about 720 per dm2. Ed.]) show that a large scale introduction of such honey-combs represents an important reserve for the increase of the bee-hive's productivity in all sectors. [Average yield increase 11.1-16.9%] To this aim it is necessary that the honey-comb should be build [sic] up first - during intense harvesting - in other colonies or in the respective colonies for honey-storage, and only afterwards it should be used for brood breeding."
On the other hand Marcel Arnst, a later compatriot of Baudoux, wrote in 1996:
"I, myself, together with other beekeepers, often compared colonies on 750 cell foundation at a comb distance of 37 mm with colonies on natural built comb at a distance of 34 mm. The 'natural' colonies were always stronger, developed faster and had less winter loss. As a result, they gave more honey (an average of 20% in my apiary of 30 beehives). The bees were also healthier: This past year there was an outbreak of chalkbrood, only the 'natural colonies' had no trace. If foundation with the natural number of cells (±850) was available, I would fit all my hives with it."
fig.1. test comb showing comb based on 'British worker' (850), 'Drone' (550) and 'Continental worker' (700) foundation built out by the bees of a single colony. photo: b.a. cooper

BIBBA experiments

From the early 1960s Beowulf Cooper and other BIBBA members began to look at the question of cell size from a breeding rather than a management point of view. They proposed that each geographical race or subspecies of Apis mellifera had its own optimum cell size or range of cell sizes, related to its genetically determined body size. As is well known, the different races are measurably different in size as well as in body proportions, and these differences are often readily detectable with the naked eye. Thus Ruttner describes A. m. mellifera as "large, broad with short limbs"; A.m. carnica and ligustica as "medium size, long limbs". The idea was that bees of a given race would tend to do better in every way when on their own optimum cell size, while using unsuitably sized foundation would result in selection pressure in an undesirable direction. Two BIBBA groups of the day experimented with foundation of different cell sizes, as a result of which Beowulf Cooper wrote:

"When given large cell foundation of 700 cells per square decimetre instead of the 'standard' type (850 or 800) normally sold in the British Isles, native strains produce appreciably larger bees, with wing length and breadth 5-10% greater than the same strains kept on standard foundation. Italian and Carniolan strains tested respond little in this way: i.e. they appear to be full-sized on 800 or 850 comb already and some types such as the Buckfast Israeli bees of the 1960s are evidently amply stretched already. In a great many cases these bees produce drones in 700 cell comb, and they are clearly genetically smaller bees.

"While all bees appear to prefer the cell size they were brought up on, and need some coaxing to work a new foundation size, just as they may need coaxing into sections, once they have made the change all the native strains so far tested have preferred the larger foundation. When native bees reared on 700 comb are allowed to build their own comb without guidance, they build appreciably larger than 850: usually 720-750. [A very similar finding was reported by Baudoux. Ed.] The 700 size seems a good one to start with for native strains, to allow for winter shrinkage of the comb and cocoon narrowing of the cell aperture.

... "Larger body size must tend to reduce heat loss and is probably associated with larger (if more dilute) nectar loads. The smaller, faster-flying strains from hotter, drier climates may find that with more concentrated nectar it pays to make more but shorter journeys."

Along with the larger cells a frame spacing of 37 mm instead of the more usual 35 mm was recommended for native British Isles bees.

One can see from this perspective one reason why the original large cell controversy and for that matter the modern small cell controversy are indeed controversies - the protagonists of the different arguments may to some degree be working with genetically different types of bee and not comparing like with like.

At the same time one can also see how bees might tend to build similarly-sized cells to those in which they were themselves reared, independently or partly so of their genetic makeup. If bees
     a) are affected by the size of the cells in which they were reared; and
     b) use parts of their own bodies to gauge the size of the cells they are building (which seems to be the case), then necessarily that size will bear a relationship to the size of the cells built by the previous generation. Thus what might seem to be a genetic effect is in fact an environmental one. The bees are not genetically any different and yet they can pass on an acquired character to the next generation in a pseudo-Lamarckian way.

This has the important corollary that the "natural" cell size found in a colony which has been allowed to build its own comb (not from foundation) may actually be the size of the foundation used by the beekeeper from which the colony swarmed or was otherwise derived. Possibly this effect could persist over many generations, perhaps indefinitely.


One merit of the small-cell controversy of recent years, to be discussed below, has been to concentrate people's minds on how cell size is actually measured. I started to write this article on "World Measurement Day". An article by Clive Cookson in the Financial Times that day mentioned some disasters caused by discrepancies in measurement. A recent NASA Mars orbiter crashed at a cost of $125m because one team of computer programmers was using feet and inches, another metres. Nearer home, privatised rail companies had to shave the edges of station platforms at great expense after rolling stock had been ordered 5cm too wide.

Measurement of foundation cell size has been carried out by so many different methods over the years, often hard to relate to each other, that one suspects some of the differences in opinions could be partly the result of different mensuration systems. Cells per linear inch; cells per linear centimetre (both measured in various directions, with or without cell walls); shortest distance across one cell; square centimetres per 100 cells; cells per square inch or per square decimetre on one or both sides of the comb (with the "square" being variously defined and shaped) - all these and more have been used.

Useful conversion tables between many methods were published on the internet by Allen Dick. One whole table is devoted to conversion between four different methods of calculating cells per square decimetre: The "rhombic (old world) method"; the "US Department of Agriculture method"; the "Baudoux method" and the "Rietsche method". These four methods will give different measurements varying by 15% or more between the highest and the lowest, for the same piece of comb.

Because of this confusion the protagonists of the regression to small cells advocate the system of "measuring across ten contiguous cells", expressed in centimetres to two decimal places. (Alternatively it can be regarded as the measurement of a single cell expressed in millimetres.) (See Fig.2). Worker cell measurement can come up with a figure ranging from about 4.50 to 6.30. The "small-cellists" prefer a figure of about 4.9, whereas "typical" foundation of 850 cells per dm2 (on all methods except the "rhombic") would be the equivalent of roughly 5.2, and 700 cells per dm2 about 5.7.

When I started to measure foundation I was surprised to find that the horizontal width across the cell is always greater than the two diagonal distances (measured in both cases between parallel walls). Is this to allow for vertical stretching of the comb in use? Anyway, to take account of this the system uses an average of three measurements at 60°, usually between the centres of parallel walls. Thus the measurement is averaged over 29 or 30 cells.

fig.2. modern system of averaging the distances between the centres of parallel walls across ten contiguous cells in three directions. diagram after spivak et al., reproduced in lusby.

This system allows for relatively easy measurement in the field or laboratory without requiring a high degree of skill. Dave Cushman writing in Bee Improvement 9 advocates the universal adoption of the system by suppliers of comb foundation, with the relevant figure defined as the "cell pitch" to be stated by the supplier in all cases. One or two manufacturers have started to do this, but beware - checking reveals that the actual pitch is not always exactly as stated!

Fig.2. Modern system of averaging the distances between the centres of parallel walls across ten contiguous cells in three directions. Diagram after Spivak et al., reproduced in Lusby.

The small cell controversy

Since the mid-1990s the South Arizona commercial beekeepers Ed and (particularly) Dee Lusby have led a vociferous campaign against what they see as a long-term trend towards large cells, and for a controlled "regression" to a small cell with a pitch ideally of 4.9. If Baudoux's primary aim was to boost honey production, and BIBBA's was to suit cell size to race and strain of bee, then it is fair to say that the Lusbys' is to promote a "biological beekeeping" with bee health as its focus. This can be seen as part of the wider "green" or "organic" agenda.

Bee health recently has of course been dominated by the problem of varroa and, in the USA, that of tracheal (acarine) mites, and the Lusbys do indeed claim that bees reared on small cells, among many other virtues, show increased resistance to these pests. Advocacy of small cells has even reached the pages of Bee Improvement (No. 7) in the shape of an article by Chris Slade, and some suppliers have started to sell what they (sometimes inaccurately) denominate as 4.9 foundation.

Although the Lusbys keep large numbers of colonies and claim success in combating varroa with the aid of the small-cell method, scientific attempts to validate their claims have had mixed results, from mildly encouraging to negative.

A group of Brazilian researchers (De Jong et al.) found that:

"When a piece of European bee comb is implanted within an Africanised bee comb, the larger diameter cells of the European comb attract more mites, even though the larvae in the two cell types come from the same queen"

Another group of Brazilians concluded (Apidologie 26 (1995), p.381-386) that with both Africanised and Italian bees, infestation was greater in large cells, but that this was not the main reason for the apparently greater resistance in the Africanised bees.

Eric Erickson and others state that

... "reduced cell diameter may have a limited impact on varroa ... and population dynamics, and on colony performance."

S.J. Martin and P. Kryger working with the African honey bee A. m. scutellata concluded that:
"Although reproduction of Varroa sp. is affected by the space between the developing bee and cell wall, reducing cell sizes as a mite control method will probably fail to be effective since the bees are likely to respond by rearing correspondingly smaller bees which explains the close correlation between cell and bee size."
Swedish researcher Mia Davidsson studied varroa reproduction in cells of 640, 770 and 900 cells per dm2 and states:
"The results do not indicate that the reproduction of mites was substantially influenced by cell size on worker bee brood."
At this point it may be worth mentioning experiments reported anonymously in the Beekeepers News by an unnamed German researcher who produced a plastic comb with cells wider at the bottom than at the mouth. The fact that the larva's body never touched the cell walls confused the varroa which was unable to reproduce. All this was observed through Perspex-backed cells. Though the details as reported at third-hand are confusing, this would seem to suggest an anti-varroa effect of large-sized cells.

The Lusbys do pay some regard to geographical variations, and have produced a world map of temperature zones, which they correlate with cell sizes, 5.2 being the biggest recommended, for cool temperate areas. The scientific basis for this correlation is not stated, however, and it is not related to Apis subspecies. Some of their own bees have been morphometrically tested and found to be somewhat Africanised, others a mixture of Carniolan and Caucasian. The climate of Arizona is certainly much drier and more extreme than anywhere in northwest Europe.

The way ahead

I have suggested that the different opinions which have surfaced over the years have resulted partly from the differing objectives of the various parties, and partly perhaps from the confusion in measuring systems. We should be grateful to the Baudoux camp for drawing attention to the whole question of cell size and its importance; to the Cooper camp for reminding us that what works for one bee subspecies may not work for another; and to the Lusby camp for sorting out the measurement question and raising the question of disease resistance. Nothing will be gained from defending entrenched positions.

A few years ago I decided to experiment with large cell foundation, unaware at the time of the small cell controversy. I used an old "Leaf" foundation press, no longer manufactured, which I measured at 5.90 (about 663 cells per dm2: 33% larger than the smallest South African sample). Learning to produce my own foundation was interesting and fun, though very time-consuming. My bees, which I would describe as quasi-native, drew it out with alacrity and the queens filled the combs readily with worker brood. A few colonies tended subsequently to convert some of the cells to drone if they did not have much other drone comb in the hive. Varroa infestation has decreased - I don't see that as necessarily attributable to cell size, but nor does it support the view that large cells encourage varroa.

There seem to be plenty of beekeepers experimenting with small cells, and of course most people will continue to use medium-sized foundation as long as it is the standard type on offer. In the spirit of experimentation I therefore intend to carry on with large cells for native or near-native bees.

If a regular demand could be established, perhaps a commercial source of large cell foundation could be located or some manufacturer could be persuaded to recommission or manufacture the necessary dies for factory or faster home production of foundation. I would like to hear from any other beekeepers who would be interested in large cell experiments.

Philip Denwood

fig.3. test comb showing comb based on (950), and (700) foundation. photo: b.a. cooper Fig.3. Test comb showing comb based on (950), and (700) foundation. Photo: B.A. Cooper


Anon. "A varroa treatment without chemicals", Beekeepers News 53 (October 1991), p.10.

Cooper, Beowulf A. The Honeybees of the British Isles. BIBBA 1986.

Cushman, Dave. "Foundation cell size", Bee Improvement and Conservation9 (Spring 2001) p.17.

Davidsson, Mia. "The influence of cell size in Varroa reproduction." http://www.algonet.se/beeman/research/cell.htm

De Jong, D., D. Message & M. Issa "The influence of cell size on infestation rates by the mite Varroa Jacobsoni." XXX International Apicultural Congress,Apimondia 1985.

Dick, Allen. http://www.honeybeeworld.com/misc/cellcount.htm

Erickson, E.H. Jr., Gary Richardson, Kevin Kehl, Dennis Arp & Brett Cameron. "Effects of comb cell diameter on parasitic mite infestations in honey bee colonies." Reported on http://www.beesource.com

Grout, R.A. The Hive and the Honey Bee. Dadant, Hamilton Il., 1973. p.235ff.

Lusby, E. & D. http://beesource.com/POV/Lusby

Martin, Stephen J. & Per Kryger. "Reproduction of Varroa destructor in South African honey bees: does cell space influence Varroa male survivorship?",Apidologie 33 (2002), p.51-61.

Root, A.I. The ABC and XYZ of Bee Culture. A.I. Root Co., Medina Ohio. 1972. p.134ff.

Ruttner, Friedrich. Breeding Techniques and Selection for Breeding of the Honeybee. BIBBA 1998.

Slade, Chris. "Size matters: a discussion on worker comb size preferences", Bee Improvement and Conservation 7 (Autumn 2000), p.27.

I can make an additional comment... The "Leaf" foundation press mentioned above was one of a batch of ten manufactured especially for the project concerned. They were made by spraying sheets of 640 foundation with a conducting medium and then a coating of copper was electrolytically deposited to produce the facings of the plates. The original invoice for the ten sheets of 640 foundation, that were used to make the moulds, is in the possession of Brian Cramp. Leaf presses were made as standard in 5.45 mm cellsize.